Run the Vibe Innovation Framework red team protocol on any artifact (problem statement, concept, business model, experiment design, prototype, decision). Use when stress-testing an artifact against adversarial framings outside the usual phase Step 7 moment, or as a standalone sanity check before locking an output.
Scanned 9/12/2026
Install to Claude Code
npx -y skills add aibot88/sec_skill_store --skill innovate-redteam --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Innovate Redteam?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/aibot88-innovate-redteam)More formats (shields.io, HTML) on the badges page.
---
name: innovate-redteam
description: Run the Vibe Innovation Framework red team protocol on any artifact (problem statement, concept, business model, experiment design, prototype, decision). Use when stress-testing an artifact against adversarial framings outside the usual phase Step 7 moment, or as a standalone sanity check before locking an output.
user-invocable: true
argument-hint: "[phase-N section-X | free | decision]"
---
**Target requested:** $ARGUMENTS
Read the following files in order before acting.
1. `.claude/docs/red_team_protocol.md` (the canonical protocol, with question sets per phase and per lens)
2. `.claude/docs/principles_and_antipatterns.md` (the cognitive bias watchlist that the protocol implements)
3. The relevant phase prompt only if a Phase artifact is targeted (for example, `phase_2_ideation.md` for Phase 2 Step 7).
Then execute the protocol exactly as specified there. The protocol is self-contained: triggers, lens catalog, question sets per phase, scoring rubric, output template.
## Operating notes
1. **Input resolution.** Parse the argument into one of three input shapes. `phase-N section-X` resolves to the corresponding ICD section and runs the phase-specific question set. `free` plus a free-text artifact runs the generic set. `decision` plus a pending decision runs the Phase 5 decision question set. If the argument is empty, ask which input shape applies before proceeding.
2. **Lens count.** Use at least three lenses, with at least one user-side, one competitor-side, and one failure-mode lens. Two passes are required if the first pass produces only *Addressed* and *Out of scope* outcomes.
3. **No straw men.** Each lens generates the strongest version of its challenge. Weak challenges are filtered out before presenting to the user.
4. **Logging.** For phase Step 7 invocations, the summary lives inside the phase output. For standalone invocations, log under Section 8 (Decision log) as a Strategic entry with type `Red team`.
5. **Working language.** Detect the user's working language from their last message. Translate the question sets and the summary block into that language. Keep the structure and tone faithful.
6. **Routing.** End with the routing line: proceed to the next step, iterate within the current step, trigger a mini-gate loop-back (use `/innovate-loopback`), or escalate to a kill or pivot decision (use `/innovate-phase 5`).
Respond in the language the user is writing in.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!